[Tagging] Feature Proposal - RFC - boundary=forest(_compartment) relations

[stevea]

Again, +1 to everything that Kevin says:  he's spot on that "these things are complicated, yet we can and do tag sensibly."

Though, (switching back to my voice, not Kevin's) "not always."  That's due to disagreements in definitions (largely, in my opinion).  A tag, set of tags or tagging scheme, with enough discussion ideally including wide, across-Earth consideration (not easy, certainly not always done) can USUALLY be hammered out, with OSM "tagging well" after that.  However.

"We" (virtually everybody on earth) means something slightly different with the word "forest" (and how a forest might be or is used in the region / jurisdiction in which somebody is speaking).  It may be strong language to say it, but unless you know this about how people actually talk about forestry, unless you have a repeated sense of how confused is this word, you could agree many, even most, are properly-stated "ignorant of the complexities."  I don't say that to be mean or insulting about it, I say that because it "forest" is a wildly confused topic where even agreeing on what the word or associated terms about it are (let alone how far it extends or smears into other contexts) remain constant sources of confusion, disagreement and debate.

I, as an OSM volunteer, wish this gets better.  I would think it starts with agreement on definition of terms, though once that starts, (often) so do disagreements.  Still, I climb the "forest" mountain with others, partly right here and like this.

I recognize that in some, even many places on Earth, "forests," as they overlap with landuse as they overlap with landcover as they overlap with ownership ... on and on this goes ... are "tagged accordingly."  (Using a Keep It Simple Silly premise).  We have tags for many aspects of these, which in some cases (e.g. ownership), data have some "standardized" (widely agreed, implemented, documented) tagging while other data have more-local, a-bit-odd tagging.  We have tags for the legal underpinnings of what a particular "land" (its polygon) derives from in the eyes of the law in that part of the world.  I think these are not 100% harmonious, though due to our plastic tagging, people have coined sensible keys for these.  In many cases, tags are well-documented in our wiki.  (Forest is documented in our wiki, we essentially say "this is widely confused").

I don't want to state what is sorely obvious, but I do believe "tag your best" still applies.  Sometimes only a whisper of fact is known ("this is a forest, or so I've heard, or so people around here call it, that's about all I know..."), sometimes much more detail is not only known, but is required for real-world use (well, CAN I canoe, fish, hunt, shoot, camp, burn downed wood in a campfire... here?).  There is a lot OSM wants to do in this tagging pipeline:  say "there is a forest here," as well as answer all those and potentially many other questions.  Really, it's a multitude of tagging pipelines, as we're not sure what a data consumer in the future will look for, infer, or actually be able to say "yup, OSM answers the particular, exact question I have years-into-the-future about what I can do or want to know about 'right here' with specificity, compared to the years-ago when these data were written by a volunteer Contributor who knew 'this much, on that day.'"  That's the power of, and the potential for confusion of, this syntax we call "tagging."

We stand tall as we do that, we get better as we grow, but there WILL be failures when somebody reaches for a particular geographic answer and OSM is unable to provide it.  That simply comes with the territory, since no database can be all answers to everybody forever.  Can we (do we) successfully answer a great many geographic questions, every minute of every day?  Yes, OSM does.  Let's keep up that good work by writing the right data for those questions to be answered.  We do so with well tagged, crisp syntax, minimal- or no-ambiguity key-value pairs on our data.  That's what OSM is and does, we are millions who author geo data.  We do well here, though we certainly can improve.  Some of the most important things we do (in OSM, on this tagging list) are disambiguate definitions and iron out misunderstandings.  These are not easy tasks.  Growth comes, yet slowly and with much (needed) deliberation.  Sometimes, there are regional conventions and ways of mapping and tagging which are simply "how things are done in a few disparate places in the world, yet they unify together with a tagging scheme like this."  We wish for the map to be its best for us as we tag our best and create tagging as our best.

Map well, tag well, discuss well.  Even when it is difficult.  Thank you for reading.